Author Topic: Tesla Coil Q's  (Read 14612 times)

0 Members and 1 Guest are viewing this topic.

Offline @rtTopic starter

  • Super Contributor
  • ***
  • Posts: 1051
Tesla Coil Q's
« on: September 21, 2015, 03:07:48 pm »
Hi Guys :)
I’m so sorry.. yes, the cliche Tesla coil project that I never got round to because I was doing projects with merit :D
Admit that it’s so sexy you want to lick it though! ;)

I have a few questions because the TC specific forum I joined, the registrations are not checked regularly.

1) This couldn’t have worked to power the whole world. How is it different than any RF carrier wave?
     I imagine it better to use a torus and do something local than try to charge the ionosphere.
2) How is it legal today? I am previously licensed amateur operator, and wasn’t allowed to make this sort of noise,
    or even make my own transmitter with F call. I imagine these make noise on harmonics at least into VHF.
3) Solid state units, it appears, are driven with a gutsy H-bridge of some variety where turns ratio applies.
    Why not operate in four cycles instead of two?
    ie. drive primary coil +/-, turn off and allow secondary collapse, drive coil -/+, turn off and allow secondary collapse.
4) Does anyone have a 15kV neon former going cheap for this guy’s big Brother?

Cheers, Brek.

 

Offline eneuro

  • Super Contributor
  • ***
  • Posts: 1528
  • Country: 00
Re: Tesla Coil Q's
« Reply #1 on: September 21, 2015, 10:21:49 pm »
3) Solid state units, it appears, are driven with a gutsy H-bridge of some variety where turns ratio applies.
Trying to find spare time for Tesla Coil, but it will have to wait a few months, I guess ;)

Anyway, I've simulated sample ZVS driver and I've almost soldered it, but didn't tested, yet, however, I've made also teardown of CRT monitor flyback, so It looks like I've part of circuit which hopefully could work like this below?



Decent lighting-maybe I'll join Tesla Coil elite club soon, but of course some research needed to keep things safe, but I've RF pilot to turn such Tesla Coil from.... Faraday cage   :-DD
12oV4dWZCAia7vXBzQzBF9wAt1U3JWZkpk
“Let the future tell the truth, and evaluate each one according to his work and accomplishments. The present is theirs; the future, for which I have really worked, is mine”  - Nikola Tesla
-||-|-
 

Offline tomlut

  • Regular Contributor
  • *
  • Posts: 57
  • Country: au
Re: Tesla Coil Q's
« Reply #2 on: September 22, 2015, 10:19:14 am »
1) It's not different. It is still subject to the distance square law. Even allowing for using the ionosphere  to earth distance as a waveguide, transmitting large amounts of power efficiently is not feasible. Tesla eventually realised this according to a documentary I watched recently.

2) Today's coiler's are after long sparks not RF radiation. Still if someone complains and you are caught (unlikely), depending on your attitude you'll probably just get a warning and a cease and desist notice.

3) IGBT bricks in a full H bridge are the go. There's a great book on it http://www.lulu.com/shop/daniel-mccauley/drsstc-building-the-modern-day-tesla-coil/paperback/product-343866.html

4) You will have to pry mine from my cold dead hands.

Tip: Next time you design a coil don't go much above a 4:1 aspect ratio for the secondary coil. It has been proven to be the sweet spot (your coil looks a bit tall). Also when you move to higher power build a flat pancake primary. You will get too much coupling, causing racing sparks along the secondary if you even raise the pancake coil up in a cone. I had to rebuild this one (see big coil pic):



A few pics of my gear:

Controller:


Mini coil 6kV 30mA NST powered. Aspect ratio of secondary is too long and the top load is too small.


Big coil 11kV 80mA NST powered. Need a bigger top load but this performs very well as is (note the flat primary).


Same 11kV NST powering a 5 ft Jacobs Ladder
« Last Edit: September 22, 2015, 10:26:17 am by tomlut »
No I won't fix your TV for a carton of beer.
 

Offline tomlut

  • Regular Contributor
  • *
  • Posts: 57
  • Country: au
Re: Tesla Coil Q's
« Reply #3 on: September 22, 2015, 10:33:03 am »
Also bit tip: Hammer a bunch of your own earth stakes into the ground in a star pattern and hook them together with heavy duty wire, copper tape or braid. Do not use a building safety earth as your coil earth. It is not an adequate RF earth and you will blow stuff up in the house. Ask me how I know...

EDIT one last thing: if you really want to be legal as far as the power company goes consider making an ELV isolation transformer from two arc welders: http://tesladownunder.com/Misc.htm#Australian%20safety%20requirements
« Last Edit: September 22, 2015, 10:36:57 am by tomlut »
No I won't fix your TV for a carton of beer.
 

Online Marco

  • Super Contributor
  • ***
  • Posts: 6693
  • Country: nl
Re: Tesla Coil Q's
« Reply #4 on: September 22, 2015, 11:01:22 am »
Big coil 11kV 80mA NST powered. Need a bigger top load but this performs very well as is (note the flat primary).

It seems to want your beer.
 

Offline @rtTopic starter

  • Super Contributor
  • ***
  • Posts: 1051
Re: Tesla Coil Q's
« Reply #5 on: September 22, 2015, 11:16:16 am »
Wow, nice one :) and nice torus. Did you spin that yourself?
and thanks for the advice :)

The small one, the objective was to tune it to an even frequency to drive it
with a circuit based on a crystal clock.
It’s 2MHz which is easy to derive from any 2/4/8/16/20MHz crystal and digital logic.
« Last Edit: September 22, 2015, 11:18:42 am by @rt »
 

Online T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21606
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Tesla Coil Q's
« Reply #6 on: September 22, 2015, 11:36:48 am »
Ah, is this what it is... (specifics always come in handy https://www.eevblog.com/forum/projects/fast-power-fet-example/?topicseen )

Crystal control gets you nothing, because the secondary's resonant frequency is variable.  Consider: as the voltage rises, and air breaks down and becomes conductive, the top load effectively increases.  It's like connecting more capacitors in parallel with the top load, but connecting them through moderately small resistors (arcs), so it's lossy as well.

The combined effect means that resonance will drop, and Q will drop considerably, as delivered power rises.

It's worse if you want to interact with it at all.  Proximity shifts the resonant frequency, for the same reason (but with less resistance).

The 'junk' version (the original, spark gap and metal bits, no active circuitry) is fine with this, because the physics of coupled resonators allows for energy transfer even when they are out of adjustment.  It's also a very wideband signal source (air sparks have a risetime in the fractional nanosecond range), so there's no shortage of excitation if peaks shift elsewhere.

So, for a transistorized version, what's needed is a resonant tracking controller, either an oscillator (well enough coupled to the secondary that it operates at resonance), or a PLL (phase locked loop) system.  The latter is quite common, using a CD4046 PLL, and a lot of input conditioning and output drive circuitry.

And no, you can't simply wire up an oscillator (like the classic ZVS circuit) and add gate driver chips and bigger transistors: this doesn't work because the circuit must have a linear startup mode, which gate drivers are specifically made to avoid.  If the circuit can't find the center resonance right away, it'll start up in some bizarre parasitic, chaotic mode and deliver a lot of heat and no power.

As a plus, the basic controller is also what's needed for induction heating, for the same reasons.  An induction coil's inductance and Q varies strongly with load, and in general, cannot be driven from a constant frequency.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline @rtTopic starter

  • Super Contributor
  • ***
  • Posts: 1051
Re: Tesla Coil Q's
« Reply #7 on: September 22, 2015, 12:01:23 pm »
I have just discovered that prior to reading this. It’s actually 1.676 MHz that it settles at (I was getting another small occasional peak interfering with scope measurement).
I have a Yaesu FRG-7 that tunes it at 1.6 MHz to verify, but it moves in and out as I move anything including my hand near it.

That’s not the only reason for the FET question (just the reason for that speed).
I have found rocks that contain iron and stick to magnets, and trying to make a rock core transformer.
« Last Edit: September 22, 2015, 12:04:07 pm by @rt »
 

Offline Psi

  • Super Contributor
  • ***
  • Posts: 9889
  • Country: nz
Re: Tesla Coil Q's
« Reply #8 on: September 22, 2015, 12:09:15 pm »
Here's mine

Greek letter 'Psi' (not Pounds per Square Inch)
 

Offline @rtTopic starter

  • Super Contributor
  • ***
  • Posts: 1051
Re: Tesla Coil Q's
« Reply #9 on: September 22, 2015, 02:19:34 pm »
Nice :)
The fact that interfering with the oscillator with one’s hand moves it’s frequency is quite helpful.
This should also be achieved if one were able to elevate the torus without touching it?

I have completed one tuned LC receiver circuit that can light the LED from further away,
and can also tune the carrier in and out. So if I duplicate this one I should just have to tune them slightly apart,
and control the two LEDs by waving my near the torus.

It might be working on an harmonic, I haven’t tried ringing it to show it’s the fundamental frequency,
but I would guess that it is.

« Last Edit: September 22, 2015, 02:21:29 pm by @rt »
 

Offline eneuro

  • Super Contributor
  • ***
  • Posts: 1528
  • Country: 00
Re: Tesla Coil Q's
« Reply #10 on: September 22, 2015, 06:59:21 pm »
Here's mine
Wow, 1 meter high or so?  ;)

Iteresting tricoil here:


Anyway, what is benefit of using toroid on top instead of eg. large sphere aluminium ball at the same surface (area)?
It is easier make toroid or something else...?  :-\
I've some idea howto make huge 1m in diameter aluminum foil based sphere, so any advantage in using toroids instead of cpnductive balls (spheres)?  :-/O

I've drilled my ZVS driver, so maybe with salvaged and modified CRT monitor flyback (i've added splited a few turns on its primary) I will have working a few kV HV DC pulses power source (it has built in HV diode)....

Update: Never mind, I've found a few usefull links ;)
http://electronics.stackexchange.com/questions/168997/why-is-a-toroid-most-often-used-as-the-top-load-on-a-tesla-coil

The Zeus Tesla Coil
http://www.hazardousphysics.com/main/zeus/The_Zeus_Tesla_Coil_3.html

The Tesla Coil Bible (PDF)
https://my.pcloud.com/publink/show?code=XZMh2VZfoGLajbllSVE3Us4g8el17BbXQ9y

Serious business  >:D



« Last Edit: September 22, 2015, 08:38:06 pm by eneuro »
12oV4dWZCAia7vXBzQzBF9wAt1U3JWZkpk
“Let the future tell the truth, and evaluate each one according to his work and accomplishments. The present is theirs; the future, for which I have really worked, is mine”  - Nikola Tesla
-||-|-
 

Offline @rtTopic starter

  • Super Contributor
  • ***
  • Posts: 1051
Re: Tesla Coil Q's
« Reply #11 on: September 23, 2015, 03:18:28 am »
I would guess it’s about radiation pattern.
There was a difference when I replaced a flat circle with a torus on my little one.
 

Offline eneuro

  • Super Contributor
  • ***
  • Posts: 1528
  • Country: 00
Re: Tesla Coil Q's
« Reply #12 on: September 23, 2015, 01:21:48 pm »
There was a difference when I replaced a flat circle with a torus on my little one.
Flat circle had still sharp edges, since sphere has curvature with the same surface radius over whole its surface, so as I read linked above posts about this, it makes discharges more unpredictable, but I'm, not looking for huge discharges and this HV stull will be closed inside grounded to different earth with cage itself, so I think I will stick with sphere made of many spot welded aluminium triangles based on https://en.wikipedia.org/wiki/Icosahedron divided to more tirangles  ;)

I will experiment first with HV flyback and I will try multiply its output HV using one of those Charge Pumps: An Overview and I'll test Dickson charge pump.

HV Sphere wil be discharged to another sphere around it -grounded to create seperate earth faraday cage, to ensure that this HV stuff can be put in the garden and operate maybe for longer time than Tesla Coil while I expect much less power consumption and not too many spark gaps  needed (maybe to limit flyback output voltage to 10kV o so)  :popcorn:
I've soldered prototype ZVS flyback driver like this:



This is not high power version, but just as proove of concept simulated in circuit sumilator to see if it works with salvaged CRT monitor flyback in real circuit.
This double layer PCB require a few corrections to help easier solder diodes on the bottom, but it is very close to send to manufacture plenty of them at very low cost if it works  8)
It is 5cm x 2.5cm PCB and additional three connectors 9slightly smaller than those 5mm to connect car starter battery(ies) ) will be soldered on the bottom to connect directly to custom split fly back primary...  :-/O
12oV4dWZCAia7vXBzQzBF9wAt1U3JWZkpk
“Let the future tell the truth, and evaluate each one according to his work and accomplishments. The present is theirs; the future, for which I have really worked, is mine”  - Nikola Tesla
-||-|-
 

Offline @rtTopic starter

  • Super Contributor
  • ***
  • Posts: 1051
Re: Tesla Coil Q's
« Reply #13 on: September 23, 2015, 01:30:01 pm »
Hi, I have already heard of them driven with flyback transformers. or if not, a plasma speaker or similar.
I suppose there’s no shortage of CRTs, plenty of them on the footpath at collection time.

I don’t understand how the steam can get through those pipes, but I believe you sir!
This is low power.. note 1/2 Watt resistors and power FETs for soft power switch
all to drive the big scary 2N3904.



and thought I was going to have to go extreme on a counterpoise, but this works,
so I guess that’s that (made for various attachments).


« Last Edit: September 23, 2015, 02:12:18 pm by @rt »
 

Offline eneuro

  • Super Contributor
  • ***
  • Posts: 1528
  • Country: 00
Re: Tesla Coil Q's
« Reply #14 on: September 23, 2015, 06:33:24 pm »
all to drive the big scary 2N3904.
I have no idea what are you want to do, but checked this monster
2N3904 - Small signal NPN transistor and it has 200mA Ic current in absolute maximum ratings, so why softstart for maximum 12W power in this thing @ 60V?  ;)

I'll make plasma speaker first before jumpimg to really HV stuff like 1MV (10kV multiplied 100x) in TC, while probably we need something like this for 1 meter long discharges?  >:D

Completed my flyback ZVS driver and we'are ready to make some... plasma, I hope  >:D



Vertical electrolyte capacitor will be changed to horizontal, since there is enougth space on ground side of top PCB and this design will look even more compact.
Additional caps mounted on top of mosfet-can be screwed on top of mosfets drains, so easy to tune and  :-/O resonant frequency of this thing.



I decided to use oryginal flyback wire for its primary, so it looks a little bit messy now, since I had to bent bottom connectors to nicelly fit into those short splited custom flyback primary, but that is fine it is prototype.

Another monitor flyback is waiting for similar tunning if it works  ;)
« Last Edit: September 25, 2015, 12:55:48 pm by eneuro »
12oV4dWZCAia7vXBzQzBF9wAt1U3JWZkpk
“Let the future tell the truth, and evaluate each one according to his work and accomplishments. The present is theirs; the future, for which I have really worked, is mine”  - Nikola Tesla
-||-|-
 

Offline @rtTopic starter

  • Super Contributor
  • ***
  • Posts: 1051
Re: Tesla Coil Q's
« Reply #15 on: September 24, 2015, 04:38:24 am »
Ah, the board is split so I can break the soft power off... It won’t always be a 2N3904.
MJE3055T also works, but I’m looking for better.
I’ve also got some material for my next coil but a neon transformer I would really like, and hard to find.
 

Offline tomlut

  • Regular Contributor
  • *
  • Posts: 57
  • Country: au
Re: Tesla Coil Q's
« Reply #16 on: September 24, 2015, 10:52:39 am »
Wow, nice one :) and nice torus. Did you spin that yourself?

Nope. The big one is a steel doughnut that I welded a centre plate in and nickel plated. They're supplied to be cut up and used for bends in 4" truck exhausts.

I bought the little one online  from http://www.rossengineeringcorp.com/products/isolation/toroids-spheres-corona-nuts.html
« Last Edit: September 24, 2015, 10:55:15 am by tomlut »
No I won't fix your TV for a carton of beer.
 

Offline tomlut

  • Regular Contributor
  • *
  • Posts: 57
  • Country: au
Re: Tesla Coil Q's
« Reply #17 on: September 24, 2015, 11:01:13 am »
Big coil 11kV 80mA NST powered. Need a bigger top load but this performs very well as is (note the flat primary).

It seems to want your beer.

That's not me. But yeah, we noticed that after downloading the photos. It's trick of perspective. The arc is a fair way in the background. There's an earth wire hanging down from the stick half way between my mate and the coil. It would have power arced to that if it was heading for him.
No I won't fix your TV for a carton of beer.
 

Offline tomlut

  • Regular Contributor
  • *
  • Posts: 57
  • Country: au
Re: Tesla Coil Q's
« Reply #18 on: September 24, 2015, 11:19:17 am »
...Anyway, what is benefit of using toroid on top instead of eg. large sphere aluminium ball at the same surface (area)?
It is easier make toroid or something else...?  :-\

Yep a as already explained a sphere will arc equally in all directions, increasing the possibility that the secondary coil could be struck and damaged. A toroid will actually shield the secondary slightly as shown in this electrostatic plot (blue areas = higher gradient):



This flattened cardioid is also a good choice:

No I won't fix your TV for a carton of beer.
 

Offline eneuro

  • Super Contributor
  • ***
  • Posts: 1528
  • Country: 00
Re: Tesla Coil Q's
« Reply #19 on: September 24, 2015, 04:58:48 pm »
A toroid will actually shield the secondary slightly
Thanks, for those nice images ;)

I'de like to put such HV stuff in the garden, so looking for extreme high voltages but at low current, and since it should be quite easy estimate capacitance of aluminium ball at given diameter (lets say 1 meter),
hyperphysics: Isolated Sphere Capacitor


where


hyperphysics: Potential: Charged Conducting Sphere

Some tables Dielectric Constants at 20°C.

After  :-DMM it gives me for 1 meter sphere 111 pF capacity.

Update: Ups, this was calculateted for 2m in diamater sphere, since I'me made mistake and substitited radius with diameter :palm:
Capacitance of charged sphere at given diameter will be: R:=D/2
Cd(D):= 4*pi*e0*D/2= 2*pi*e0
Cd(D) ~ 56 pF *D

So, for 1m (in diameter) sphere we have: 56 pF capacitance.
I've found 10kV 100pF capacitors and 16kV 20ma diodes, so probably will try build Dickson voltage multiplier using such components with flyback output current limited by two 10kV 10MOhm resistors.
Quick estimation of energy in 100kV dickson made of 10kV capacitors gave me estimated amount of stored energy at 100kV based on 10kV VDC PSU ~275mJ using 100pF caps in each of ten stages which is not so high when we compare with 500mJ of charged car, etc  8)
It looks like enrgy in Disckson voltage multiplier, when 10kV caps in series will be used in next stages up to 100kV we have formula for amount of energy in each stage:
Ek=E0*k, where E0=1/2*C0*V0^2, C0=100pF, V0=10kV, so E0= 5mJ, when 100pF capacitor will be charged to 10kV ....
End of update -time to order elements and build this thing  >:D


What could be capacitance of those touruses you use in your TC's?  :-\

When we have capacitance and know voltage than we could estimate amount of energy and try to determine maybe from some regulations for HV stuff if such capacitance at given HV level is safe if discharge not repeted more than lets say 1Hz, however it ould be better have at least 100Hz discharge rate for visual effects...

111 pF @ 100kV gives ~ 0.556 J = 556 mJ of energy.

Howto estimate is it safe to touch and discharge by hand such charged sphere at not more than 1 Hz rate?  :-/O
When we assume than we discharge via 1k resistance than we get 3RC constant ~ 333 ns ~0.3ms ....

Than, when we calculate amount of charge at 100kV we have: 111pF * 100kV ~ 11.1 uQ , so dividing this charge by discharge time 0.3ms  we get average current: ~33.3 mA within less than 1ms , so less than 100mA, but magnitude higher than 1mA, but discharge time less than 1ms... Hot could affect person who touches such HV ball lifted 1 meter above ground?  :-\

Such huge charged ball looks a little bit scary... no plans  touch it without clear evidence prooved by some safety rules that it might be safe or maybe it isn't even at lower than 1 Hz rate-from time to time....

Maybe old school Van de Graaff generator powered by small wind turbines could be better option (more safe) than TC which might require much more power and provide much more power, I guess? ::)
Quote
A Van de Graaff generator is an electrostatic generator which uses a moving belt to accumulate electric charge on a hollow metal globe on the top of an insulated column, creating very high electric potentials. It produces very high voltage direct current (DC) electricity at low current levels.
Trying to figure out what could be maximum possible potential to achieve on 1 meter in diameter aluminium sphere (ball)  lifted  ~ 1 meter above ground, but in the garden, not inside building.
There are many factors which influence this air breakdown voltage, but we could asume eg. 15*C dry air...
Anyway, when we calculate capacity of such sphere, than probably easier to test this than calculate at very changing outside air atmosphere conditions, but to do this I need at least Van de Graaff generator or maybe TC could do this job too, if we protected TC seconday somehow?
What about... put whole TC inside such quite big 1 meter in diameter aluminium sphere (I mean with battery and circuit ) and leave only well insulated wire to earth to be able create potentials difference between lifted ball and earth?  :-\
« Last Edit: September 26, 2015, 10:32:23 pm by eneuro »
12oV4dWZCAia7vXBzQzBF9wAt1U3JWZkpk
“Let the future tell the truth, and evaluate each one according to his work and accomplishments. The present is theirs; the future, for which I have really worked, is mine”  - Nikola Tesla
-||-|-
 

Online T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21606
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Tesla Coil Q's
« Reply #20 on: September 25, 2015, 01:04:13 am »
A Tesla Coil is just a resonant transformer.  Power goes in, into a characteristic impedance, and goes out, dissipated into another impedance.

What the hell impedance is it?  It's open circuit, right?

Nah, it's got that big sparky air discharge there.  Which does two things:
1. Plasma is conductive.  So the approximate dimension of the discharge is effectively growing the top load by that much.
2. It's also lossy, so the additional capacitance can simply be modeled as a lossy capacitor.

Suppose a given construction has a 20pF top load and we drive it with enough power to ionize another 10pF worth of space.  Let's suppose the Q factor of that ionization is around 1 or 2.

The total Q then is around 3, for 30pF.

We have a free variable of what frequency to drive it at.  Let's pick something typical like 500kHz.

At 500kHz, 30pF is 10.6kohms.  The Q=3 suggests Rpar = 31.8kohms, then.

So we have 10.6k of reactance to cancel with an inductor at resonance, and a load of 31.8kohms to drive with a lot of power.

Suppose we're matching it to a solid state driver with 320VDC supply, full bridge, maybe 1-3kW worth.  That's an inverter load of 33-100 ohms.

By transformer action alone, we must use a turns ratio of 1:31 to 1:54.  What do we get from resonance?

Not much, actually.  The loaded Q is very low, so it's fair to say: the multiplication due to resonance is fairly small; and the Q factor required of the components is also small (the coil's unloaded Q could be a paltry 30 and you wouldn't mind).

We can factor out the transformer action and treat the LC resonant circuit as an L-match circuit (this would also be the measured impedance at the base of the coil, with respect to ground).  With a Q of 3, and driven by a voltage source, the multiplication factor is only 3, and the feedpoint impedance is Q^2 times smaller, or 3.5kohms.  Which, by the way, is a perfectly reasonable impedance to drive with a vacuum tube -- which explains why "plasma tweeter" circuits work as well as they do.

The coupling factor for critical coupling is 1/Q, or 0.33.  Pretty large, as resonant circuits go -- typical IF coils in radios are much smaller, with corresponding higher Q factors.  More reason to consider it not so much as a resonator but as a transformer with a high ratio.

So including resonance, we need a turns ratio of 3.4 to 6.

Yes, a rather low turns ratio; but how?  Well, pulling numbers out of my ass, of course.  A total ratio of 54 suggests no more than 17kV output, which isn't going to make nearly 10pF worth of lossy plasma!  20pF is a fair value for a large top load, at a power level probably more like 10kW.  The lossiness of a plasma discharge may be quite a bit lower than I guess.

But whether the initial assumptions are wrong or not, the design process is the same, and now that you know it, you can repeat it for your own purposes -- assuming you have a better number for the load, of course.

BTW, note that spark discharge types can achieve a much higher peak power, due to the capacitor discharge event.  An SSTC can't do that: it has to be rated for as much current as it has to deliver, period.

Let's see if I can salvage some of these numbers anyway!  It seems more reasonable that 1kW should be able to do 8" of sparks (CW), which might amount to 200kV and 5pF of plasma.  1kW at 200kV is 40Mohm.  5pF at 0.5MHz is 63kohm.  The Q should then be over 600... which seems unbelievably high!  (The inductance would be 20mH and the impedance at the base would be about 100 ohms, direct drivable from a solid state inverter, or just a few turns primary.  Resonant coupling would be unnecessary and undesirable, since the Q factor is so high; needless to say it's also critical to have a high Q coil, which might be impossible at this frequency.)

That still seems wrong, maybe not quite as wrong as the first try, but unrepresentative anyway.  Maybe CW sets of that power level are usually run in the MHz, so the impedance is lower?  That would help out.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline eneuro

  • Super Contributor
  • ***
  • Posts: 1528
  • Country: 00
Re: Tesla Coil Q's
« Reply #21 on: September 25, 2015, 12:28:08 pm »
A Tesla Coil is just a resonant transformer.  Power goes in, into a characteristic impedance, and goes out, dissipated into another impedance.

I've run successfully my HV DC power supply made from salvaged CRT monitor flyback transformer with custom split a few turn primary and ZVS drived designed in circuit simulator shown on photos a few posts above and I was able make decent plasma sparks  :scared:
ZVS flyback driver high voltage plasma from 12V car battery


I've used this DIY HV 0.5 meter long stick (20mm hole inside to have some scale idea) to avoid any hand touching of flyback output wires while playing with this HV plasma output ;)


However, I've no idea which is operating frequency of this thing (i didn't looked at transformer primary waveforms), but even if this is decent HV source (with a few mm plasma sparks)  probably not usable in Tesla coil, I guess, since I have ONLY HV DC output (NOT AC  :'() , since this oryginal flyback transformer has HV diodes and as its name suggests, yep it is flyback, so design optimised for flyback operation, not for full wave AC input/output  :-\

Anyway, I've made some measurements of input power when plasma arc is ignited at it draws ~2A @ 12V  (23W) car starter battery and only 120mA  (1.4W) when no plasma spark loads  8)
I had only one 100nF capacitor in this first test run after assembly of this prototype PCB and no heatsinks on IRFZ44N mosfets-those things goes warm when running for a fe seconds, so yep bigger caps bank and mosfet heatsinks are needed for longer operation with current maybe limited even more to 0.5A or so for continous operation, but it was only test of my ZVS concept at it works fine  :popcorn:

Is it possible use such HV DC (actually it must be high frequency  HV dc pulses while no significant output capcity inside this thing I guess based on example internal circuit schematics) PSU as TC input power?
My idea is, use it as input to mentioned Dickson voltage multiplier and maybe get at least 100kV DC this way, but it could be nice if I could use this PSU maybe  for small TC?  :-//

Anyone had success in powering TC with flyback transformer HV DC pulses?  :-//


« Last Edit: September 25, 2015, 12:34:59 pm by eneuro »
12oV4dWZCAia7vXBzQzBF9wAt1U3JWZkpk
“Let the future tell the truth, and evaluate each one according to his work and accomplishments. The present is theirs; the future, for which I have really worked, is mine”  - Nikola Tesla
-||-|-
 

Online T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21606
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Tesla Coil Q's
« Reply #22 on: September 26, 2015, 01:51:35 am »
It should be possible to do that; the peak power can be much higher, at the expense of a low repeat rate and duty cycle.  This is apparently what a Taser does: charge up a medium voltage supply, where the peak power from a given capacitor is useful, then discharge that capacitor into a serious step-up transformer to get the (up to 50kV?) pulse to jump the gap.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline alsetalokin4017

  • Super Contributor
  • ***
  • Posts: 2055
  • Country: us
Re: Tesla Coil Q's
« Reply #23 on: September 26, 2015, 02:57:21 am »
24 VDC from 2 12v 7AH batteries > ZVS oscillator (IRFP260n x2) > CRT flyback transformer > multi-element fixed spark gap > HV capacitor bank > primary > secondary > smooth topload from cakepan and flowervase = Hybrid ss-sg tc =




13 v in > 4046 PLL > 555 interruptor > TC4420 > IRFP260n x1 > 5 turn primary close-coupled > secondary on oatmeal box > topload made from 2 tuna cans =



Mains-powered MOT-DC two-element airblown fixed gap SGTC slideshow:





The best TC designer/calculator I've found yet:

http://www.classictesla.com/java/javatc/javatc.html


« Last Edit: September 26, 2015, 03:02:46 am by alsetalokin4017 »
The easiest person to fool is yourself. -- Richard Feynman
 

Offline eneuro

  • Super Contributor
  • ***
  • Posts: 1528
  • Country: 00
Re: Tesla Coil Q's
« Reply #24 on: September 26, 2015, 08:34:50 am »
13 v in > 4046 PLL > 555 interruptor > TC4420 > IRFP260n x1 > 5 turn primary close-coupled > secondary on oatmeal box > topload made from 2 tuna cans =
Nice sparks   :-+

I've used IRFZ44N (55VDS) with 5nF @ 12Vgs, when looked at IRFP260N datasheet and calculated the same from its total gate charge 234nC we get 4x higher gate capacitance ~20nF @ 12Vgs, so probably simple ZVS with current limiting resistors will be too slow to provide decent switching speed in the range of hundreds of kiloherts  :-\

Which is 555 frequency in this classic flyback configuration linked by you? 100kHz or more?

I've developed this simple ZVS driver, but rather as easy to make a few elements driver to make simply isolated DC-DC eg. output four galvanic insulated 24VAC from 12VDC, to be able easy power up low & high side mosfets in full H-bridge with optoisolators like TLP351F, but didn't tested it yet at hundreds of kiloherts, or so....

Anyway, meantime I've found Myth busters attempt to build Van de Graaff generator  >:D
Mythbusters Van de Graaff generator


It looks like even smaller sphere capacitor (0.5m in diamater I guess) at 100kV or so can create quite powerfull sparks when human body at given sometimes 400pF capacitance will try touch this thing and even those thearless guys strugled to touch this charged sphere  :-DD

Here I've found some usefull safety tips related to Van de Graaff generator
Van de Graaff Electrostatic Machine Safety

Quote
If you receive a "static" spark from touching a car, the energy can be as high as 500 mJ (millijoules)
...
A typical output is 30 microamps at 300,000 volts, giving a load resistance of V/I, or 10,000,000,000 ohms.
....
A healthy VDG might give 10 to 20 uA. A fast machine with a very wide belt might give as much as 300 uA.
..
When running, the VDG will produce a few tens of microamperes in the same direction as the high voltage polarity.

Howto limit output from this a few kV flyback transformer to get quite safe (providing if we do not create HV capacitamce chains on its output) curent in the rage below 1mA?
eg. by using 10kV/1mA= 10MOhm resistance ?
What about 10x   1Mohm resistors in series sealed in parafina and thermal heatshrink to create ~10cm long (HV capable?)  resistor on flyback secondary HV DC  or maybe  better look for current limit on transformer primary (or additionally)-I used light bulb in simple setup in my ZVS driver, but it will not work, since cold light bulb (21W 12V) easy conducts 3A and more, while when switched on warm nominal curent could be 1.75A, so light bulb is not an option to limit this current in ZVS driver power input, but maybe 5W has higher resistance  :-\

However, it was interesting to see, how light bulb  :box: with plasma sparks and limits is length, when light bulb warm up and input current drops, so spark is degraded, but lower current means turning light bulb off and becames cold, so input current to ZVS driver increases  making sparks higher again  8)
« Last Edit: September 26, 2015, 08:38:55 am by eneuro »
12oV4dWZCAia7vXBzQzBF9wAt1U3JWZkpk
“Let the future tell the truth, and evaluate each one according to his work and accomplishments. The present is theirs; the future, for which I have really worked, is mine”  - Nikola Tesla
-||-|-
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf